May 3, 2021

Risks of using medical story arcs

INTRO / SUMMARY

When we fit the patient's narrative into a typical medical story arc, we risk making diagnostic errors.

Watch this article as a video

The typical medical story arc

We turn our attention now to the process of converting a "patient's narrative" into a "typical medical story arc".

What is a "typical medical story arc"? I consider a typical medical story arc to be a narrative pattern of a particular diagnosis. It starts with a standard set of patient symptoms, and events that ultimately lead to a specific diagnosis and subsequent complications. If the diagnosis is considered to be the top of the story arch, many patient's who share that diagnosis has a similar presentation towards the diagnosis, and sequence of events afterwards.

In the field of diagnostic reasoning, the same idea of found in the concept of an "illness script." I'm using the terminology of a "medical story arc" as this post is part of a series on narrative language and medical documentation.

Using the story arc during the medical interview

A common diagnostic approach is to adopt a pattern recognition technique, where the physician tries to fit a particular patient's unique narrative into the narrative arc of an existing medical story they are familiar with.

The physician may question the patient to search for facts that increase or decrease the similarity of the patient's medical narrative with existing known medical stories.

Once satisfied that the patient's narrative is sufficiently fitting into a preexisting medical story arc, the physician will present the patient's story to colleagues and in documentation, not using the patient's own words and sequencing, but using the structure of a traditional medical story arc.

Risk of the medical story arc: overfitting

This approach has the risk of overfitting a patient's narrative into a pre-existing medical story. By overfitting, I mean the forcing of a patient narrative into a standard pre-existing narrative structure. This has the risk of misdiagnosis by latching onto particular aspects of a patient's narrative, and insufficiently exploring other elements of the narrative, which might have lead the physician to an alternative diagnosis.

Physicians may not be aware of all diagnoses and their associated illness scripts/medical story arcs. Furthermore, a physician's medical story arc for a particular condition may be inaccurate or incomplete.

Therefore, the process of fitting the patient's narrative into pre-existing medical story patterns at an early stage in the diagnostic process - such as during the initial interview and in the initial documentation - carries with it the risk of overlooking important facts in the history, physical, and diagnostic workup that may have led to an alternative diagnosis.

One of the biggest risks of this approach to diagnosis and documentation, is that the story will make sense. If the patient's story can be sufficiently aligned to a pre-existing medical story arc, we will think it is true because it fits. However, there is the real possibility, that the exact same story, may fit better in an alternative story arc (i.e. different diagnosis).

There is the real risk, that we may never discover, or only discover too late, that the wrong story arc was used. The reason for this, is that we like to tell stories that make sense. Once a satisfactory story arc has been identified, we write our notes as physicians in ways that include the necessary details required for the story arc we are telling. In this way, when we re-read our note, and when others read our note, we have included the information that we thought was required to tell the medical story arc. However, there is a risk that we have overlooked facts that are important because we did not realize we were overfitting the patient's story into the wrong story arc.

The gap between what the patient's narrative actually states, and what the clinical documentation presents can be at times quite large. This is, yet, again, another reason why expert physicians often prefer to re-interview patients themselves, and may often disregard what they find in the chart when dealing with complex or undiagnosed patients.

Can we survive without medical story arcs?

Physicians are lost without medical story arcs. It is such an important way that we think about the course of an illness, that we naturally start to use this reasoning process right from the start when we encounter a new patient.

However, the alternative to using stories is not much better. A common critique of medical students and junior residents is that when they present a patient's narrative, it is often a series of discrete and nonconnected facts. They present a lot of information, but they do not lead the listener (aka the attending physician) throughout the medical documentation in a cohesive story. This makes it very hard for the listener to know what facts to pay attention to, and which to potentially disregard. Therefore, medical students will be instructed and coached to present their cases so the patient's presentation fits into the arc of a known medical pattern. This makes it easier for the attending physician to ingest the story they are being told. But as discussed, it carries with it the risk of being told a story that sounds right, but is not actually the full patient narrative.

Larry Weed proposed a solution to the bias created by using medical story arcs to drive the diagnostic and documentation process. Over half a century ago, he advocated for the use of standardized patient questionnaires, standardized data collection methods, and standardized data presentation methods.

This approach minimizes the risk of overfitting the patient story to a preconceived medical narrative. He claimed that this systematic and comprehensive approach would often uncover certain aspects of the patient's health history that would have often been missed when a more focussed approach driven by searching for the facts required to fit the patient's story into an existing patient narrative is used.


Conclusion

My general impression is that a more standard approach to the history and documentation steps will more reliably lead to the correct diagnosis and avoid the errors that enter into diagnostic reasoning early on by fitting a patient's narrative into a preconceived medical story arc. However, this is quite tricky to put into practice.

We must focus on history taking approaches the encourage the patient to tell their story using their own narrative structure and documentation methods that best preserve the narrative arc and features of the patient story that they considered of importance while minimizing our natural inclination to re-order the patient's narrative into a medical story arc that we think fits better.

Subscribe for updates

Get an email update weekly, monthly, or quarterly.
Subscribe